1. Field of the Invention
The present invention relates to an image processing apparatus and, more particularly, to an image processing apparatus suitable for performing an ink jet recording.
2. Related Background Art
As an image processing apparatus for receiving multi-value image data and binarizing, there is known an image processing apparatus in which digital multi-value image data indicative of the concentrations or the like of the pixels which are input from, particularly, a scanner, television camera, or the like is binarized into the binary signals of ON and OFF and these binary signals are output to an ink jet printer.
Hitherto, in such a kind of image processing apparatus, when the image data having what is called depth information such as color images or gradation images which are sent from a host computer or the like is printed by an output device like a line printer or the like such as an ink jet printer, in general, comparison data of a fixed threshold value pattern matrix is accessed from a memory (storage device) and compared with the input image data, thereby binarizing.
Namely, such image data having the depth information is the data in which the gradations of the pixels or the like are represented by digital values as is well known and such image data cannot be directly printed by a line printer of the binary output type or the like. Therefore, the image data must be binarized before it is input to the line printer or the like. For this purpose, as a conventional method of representing the gradations by the line printer or the like, there is known an image processing method whereby a set of a plurality of dots is defined as one pixel and by determining which dots in one pixel are printed or not in accordance with the gradation level of the input image data, thereby outputting and representing the gradation (light or dark concentration of the image). However, when the image is printed by this method, there is a problem such that an unbalance occurs between the printed pixel and the other pixel and a false contour is generated, so that the printing quality deteriorates. To avoid this problem, there is used a system in which another output pattern different from the input pixels is specified and this pattern is compared with the image data having the depth information, thereby properly expressing the gradations with a wide range, that is, a binarization image processing system (including a dither system) based on the threshold value pattern comparison.
FIGS. 13A and 13B show an example of a schematic constitution of the conventional binarization image processing system in which the multi-value image data is compared with the threshold value data matrix (data) and the image is printed on the basis of the resultant binarized image data. According to this conventional system, as a threshold value matrix 2 for multi-value image data 1, for example, a pattern of a matrix of 4.times.4 as shown in these diagrams is provided. The image data 1 and matrix pattern 2 are compared by a comparator 3 and binarized. The binarized data is stored in a line memory 4. The binarized data is read out of the line memory 4 at timings as shown in FIG. 14 and output as a dot printing 5 by a line printer. In this case, since the threshold value matrix 8 has the pattern of 4.times.4, seventeen (=16+1) gradations can be expressed by an area gradation method.
By enlarging the size of the threshold value matrix 2, the finer gradations can be obtained. For example, by setting the size of the matrix to 8.times.8, 65 (=64+1) gradations can be expressed. By setting the size of the matrix to 12.times.12, the 145 (=144+1) gradations can be expressed.
FIG. 15 shows a conventional example of a circuit to perform the foregoing binarization image process. In the diagram, the input image data 1 is the digital multi-value data which is obtained by the digital values of eight bits or the like. However, since it is printed by an output device such as a dot printer, it is input to one input terminal of the (parallel) comparator 3 through a latch circuit 2. This image data is compared by the comparator 3 with the threshold value matrix data which is read out of a pattern memory 4 and which is input to the other input terminal of the comparator 3, so that it is dot developed into the binarized data of 0 and 1. When the multi-value image data is simply binarized, the gradations cannot be expressed. Therefore, when the input image data 1 is input to the comparator 3, the comparison data (threshold values) are sequentially read out of the pattern memory 4 in which threshold values (threshold value matrices) are written and the image data is compared with this comparison data, so that it is binarized to the values of 1 and 0. For example, in the case of developing one image data 1 into the pixels of 4.times.4 and binarized, the threshold value data is accessed sixteen times for the one image data 1 and successively compared with the one image data, so that the image data is binarized by the dot development of 4.times.4.
FIGS. 16A to 16D show the operations of the conventional circuit based on the pmxel development methods of 1.times.1, 2.times.2, 3.times.3, and 4.times.4. The input multi-value image data 1 is developed into the pixels in accordance with the pixel size and compared with threshold value data 8 of 4.times.4, thereby outputting the binarized concentration pattern onto a printing surface 9. "1" on the printing surface 9 shows that a dot is printed and "0" indicates that no dot is printed. In the diagrams, four kinds of pixel development methods of 1.times.1 to 4.times.4 have been shown. However, the size of the pixel is determined depending on a set of which number of ink dots by which the input multi-value image data 1 is expressed. Therefore, as compared with the image which is developed into the pixel of 1.times.1, the image which is developed into the pixels of 4.times.4 is printed as the size which is sixteen times as large as the image of 1.times.1. On the other hand, since the threshold value matrix 8 consists of 4.times.4 pixels, seventeen (including white) gradations can be expressed by the area gradation method.
FIGS. 17A to 17C show an example of the actual threshold value data of a threshold value matrix 8-1 of 4.times.4, an example of an ideal printing model which is printed by this threshold value matrix, and an example of the actual threshold value data of a threshold value matrix 8-2 of 8.times.8. Numerical values 1 to 16 in FIG. 17B represent the numbers of gradations. The threshold value data in FIG. 17C is expressed by the hexadecimal numbers. Although the threshold value matrix 8-2 of 8.times.8 in FIG. 17C is the threshold value pattern matrix corresponding to the multi-value image data 1 of eight bits, up to 65 (including white) gradations can be expressed.
In other words, according to the conventional image processing system, only 17 gradations are obtained by the matrix area of 4.times.4 and in order to further increase the number of gradations, the matrix area must be enlarged to the area of, e.g., 8.times.8 and the fine half tones cannot be expressed. However, the increase in size of the threshold value matrix causes drawbacks such that although the number of gradations increases, the resolution in output printing remarkably deteriorates and the graininess roughness becomes conspicuous. Accordingly, when the image data is the multi-value image data of eight bits, although the concentration levels of 256 gradations are provided, if the binarization is performed to execute the area modulation and the resultant binarized data is printed and output by the foregoing conventional system, in order to prevent the remarkable deterioration in resolution or graininess, for example, the gradation levels are compressed to 17 levels by using the threshold value matrix of 4.times.4 as mentioned above. In this manner, the half tone must be expressed by fairly sacrificing the gradations.